public inbox for linux-kernel@vger.kernel.org
 help / color / mirror / Atom feed
From: Andi Kleen <ak@suse.de>
To: john stultz <johnstul@us.ibm.com>
Cc: Thomas Gleixner <tglx@linutronix.de>,
	Andrew Morton <akpm@osdl.org>,
	LKML <linux-kernel@vger.kernel.org>, Ingo Molnar <mingo@elte.hu>,
	Len Brown <lenb@kernel.org>,
	Arjan van de Ven <arjan@infradead.org>,
	Roman Zippel <zippel@linux-m68k.org>
Subject: Re: [patch 13/19] GTOD: Mark TSC unusable for highres timers
Date: Fri, 10 Nov 2006 06:10:13 +0100	[thread overview]
Message-ID: <200611100610.13957.ak@suse.de> (raw)
In-Reply-To: <1163121045.836.69.camel@localhost>

	current_tsc_khz = tsc_khz;
> >  		clocksource_tsc.mult = clocksource_khz2mult(current_tsc_khz,
> >  							clocksource_tsc.shift);
> > +#ifndef CONFIG_HIGH_RES_TIMERS
> >  		/* lower the rating if we already know its unstable: */
> >  		if (check_tsc_unstable())
> >  			clocksource_tsc.rating = 0;
> > -
> > +#else
> > +		/*
> > +		 * Mark TSC unsuitable for high resolution timers. TSC has so
> > +		 * many pitfalls: frequency changes, stop in idle ...  When we
> > +		 * switch to high resolution mode we can not longer detect a
> > +		 * firmware caused frequency change, as the emulated tick uses
> > +		 * TSC as reference. This results in a circular dependency.
> > +		 * Switch only to high resolution mode, if pm_timer or such
> > +		 * is available.
> > +		 */
> > +		clocksource_tsc.rating = 50;
> > +		clocksource_tsc.is_continuous = 0;
> > +#endif
> >  		init_timer(&verify_tsc_freq_timer);
> >  		verify_tsc_freq_timer.function = verify_tsc_freq;
> >  		verify_tsc_freq_timer.expires =
> 
> 
> Hmmm. I wish this patch was unnecessary, but I don't see an easy
> solution. 

Very sad. This will make a lot of people unhappy, even to the point
where they might prefer disabling noidlehz over super slow gettimeofday. 
I assume you at least have a suitable command line option for that, right?

Can we get a summary on which systems the TSC is considered unstable?
Normally we assume if it's stable enough for gettimeofday it should
be stable enough for longer delays too.

-Andi


  reply	other threads:[~2006-11-10  5:10 UTC|newest]

Thread overview: 70+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2006-11-09 23:38 [patch 00/21] Highres / dynticks drop in replacement for 2.6.19-rc5-mm1 Thomas Gleixner
2006-11-09 23:38 ` [patch 01/19] hrtimers: state tracking Thomas Gleixner
2006-11-10  9:19   ` Arjan van de Ven
2006-11-10  9:40     ` Andrew Morton
2006-11-10  9:45       ` Thomas Gleixner
2006-11-23 22:26   ` Roman Zippel
2006-11-09 23:38 ` [patch 02/19] hrtimers: clean up callback tracking Thomas Gleixner
2006-11-10  9:20   ` Arjan van de Ven
2006-11-09 23:38 ` [patch 03/19] hrtimers: Move and add documentation Thomas Gleixner
2006-11-09 23:38 ` [patch 04/19] Add a framework to manage clock event devices Thomas Gleixner
2006-11-10  9:47   ` Arjan van de Ven
2006-11-23 22:36   ` Roman Zippel
2006-11-09 23:38 ` [patch 05/19] ACPI: Include apic.h Thomas Gleixner
2006-11-09 23:38 ` [patch 06/19] ACPI: Keep track of timer broadcast Thomas Gleixner
2006-11-10  9:51   ` Arjan van de Ven
2006-11-09 23:38 ` [patch 07/19] ACPI: Add state propagation for dynamic broadcasting Thomas Gleixner
2006-11-10  9:52   ` Arjan van de Ven
2006-11-09 23:38 ` [patch 08/19] i386: cleanup apic code Thomas Gleixner
2006-11-10 10:04   ` Arjan van de Ven
2006-11-10 10:16     ` Thomas Gleixner
2006-11-10 10:16       ` Arjan van de Ven
2006-11-09 23:38 ` [patch 09/19] i386: Convert to clock event devices Thomas Gleixner
2006-11-10 10:10   ` Arjan van de Ven
2006-11-09 23:38 ` [patch 10/19] PM_timer: allow early access and move externs to a header file Thomas Gleixner
2006-11-10 10:12   ` Arjan van de Ven
2006-11-09 23:38 ` [patch 11/19] i386: Rework local APIC calibration Thomas Gleixner
2006-11-10 10:17   ` Arjan van de Ven
2006-11-10 10:23     ` Thomas Gleixner
2006-11-10 11:10     ` Ingo Molnar
2006-11-09 23:38 ` [patch 12/19] high-res timers: core Thomas Gleixner
2006-11-10 10:26   ` Arjan van de Ven
2006-11-09 23:38 ` [patch 13/19] GTOD: Mark TSC unusable for highres timers Thomas Gleixner
2006-11-10  1:10   ` john stultz
2006-11-10  5:10     ` Andi Kleen [this message]
2006-11-10  8:10       ` Thomas Gleixner
2006-11-10  8:50         ` Andrew Morton
2006-11-10  8:57           ` Ingo Molnar
2006-11-10  9:13             ` Andrew Morton
2006-11-10  9:29               ` Andi Kleen
2006-11-11 11:14                 ` Thomas Gleixner
2006-11-11 13:51                   ` Andi Kleen
2006-11-11 13:58                     ` Thomas Gleixner
2006-11-11 13:59                       ` Andi Kleen
2006-11-11 14:08                         ` Thomas Gleixner
2006-11-10 10:35               ` Arjan van de Ven
2006-11-10 10:47                 ` Andi Kleen
2006-11-10 10:55                   ` Arjan van de Ven
2006-11-10 11:13                     ` Ingo Molnar
2006-11-10 11:28                     ` Andi Kleen
2006-11-10  9:27             ` Andi Kleen
2006-11-10 10:14             ` Alan Cox
2006-11-10 11:19               ` Ingo Molnar
2006-11-10 15:43               ` Chris Friesen
2006-11-10 11:12             ` Pavel Machek
2006-11-10 11:48               ` Ingo Molnar
2006-11-10 11:56                 ` Andi Kleen
2006-11-10 13:12                   ` Ingo Molnar
2006-11-10 12:00                 ` Pavel Machek
2006-11-10 13:14                   ` Ingo Molnar
2006-11-10 11:11         ` Pavel Machek
2006-11-10 10:28       ` Arjan van de Ven
2006-11-10 10:30         ` Andi Kleen
2006-11-10 10:37           ` Arjan van de Ven
2006-11-09 23:38 ` [patch 14/19] dynticks: core code Thomas Gleixner
2006-11-09 23:38 ` [patch 15/19] dyntick: add nohz stats to /proc/stat Thomas Gleixner
2006-11-09 23:38 ` [patch 16/19] dynticks: i386 arch code Thomas Gleixner
2006-11-09 23:38 ` [patch 17/19] dynticks: Fix nmi watchdog Thomas Gleixner
2006-11-09 23:38 ` [patch 18/19] high-res timers, dynticks: enable i386 support Thomas Gleixner
2006-11-09 23:38 ` [patch 19/19] debugging feature: timer stats Thomas Gleixner
2006-11-23 22:24 ` [patch 00/21] Highres / dynticks drop in replacement for 2.6.19-rc5-mm1 Roman Zippel

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=200611100610.13957.ak@suse.de \
    --to=ak@suse.de \
    --cc=akpm@osdl.org \
    --cc=arjan@infradead.org \
    --cc=johnstul@us.ibm.com \
    --cc=lenb@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mingo@elte.hu \
    --cc=tglx@linutronix.de \
    --cc=zippel@linux-m68k.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox